Cooperative On-Line Planning For Adaptive Map Building In Environmental Applications

Two adaptive online planning algorithms for the exploration of an unknown environment by cooperating autonomous underwater vehicles are introduced and compared. The algorithms are suited for a team of autonomous vehicles taking point-wise environmental measurements over an ocean region, with the final objective of producing an estimated map of the measured environmental quantity over the whole region. Simulative test cases in which the algorithm performances are compared with that theoretically achievable by rapidly-exploring random tree search are presented